A three-dimensional (3D) “mesh” model of a physical object can be produced by stitching together images of different sides of the object. Generating 3D models in this way presents the particular challenge that substantially all of the sides of the object need to be imaged in order to produce a complete model of the object. Failing to capture images of the object from some certain angles may result in holes, gaps, distortions, or other artifacts in the generated model (in some instances, the bottom surface of the object may be ignored if it is not relevant to the desired 3D model).
FIG. 1 shows one way of collecting images of an object 10 (e.g., a lamp) involving placing the object on a rotating platform 12 that rotates the physical object while a camera 14 remains in a fixed position. The technique shown in FIG. 1 may be appropriate for studio-like environments that provide a rotating platform, a camera stand or tripod 16, and appropriate lighting, but this may result in added cost due to the specialized equipment involved and due to the time and planning needed to bring the object to the studio environment for scanning.
FIG. 2 shows a free-form capture technique in which a user 22 captures images of a stationary object 20 using a camera 24 having an integrated screen or display. Such a camera 24 may be, for example, a standalone digital camera, a personal digital assistant, a smartphone, a tablet computer, or a custom scanner. The user 22 can walk around with the camera to image the various sides of the object 10. The set-up shown in FIG. 2 provides the freedom to scan objects in essentially any environment, without the need to bring the object to a studio and without the cost of acquiring additional specialized hardware such as a rotating platform and a camera stand.
As seen in FIG. 2, a user may encounter a usability problem when performing such a 3D scan of a stationary object 20 on a fixed platform (e.g., a table). In particular, as with shooting video, the user or camera operator 22 typically needs to watch the camera screen (or, for example, a smartphone screen if a smartphone camera is used or a camera fixture is attached to the display platform, as indicated by the dotted line in FIG. 2) in order to ensure that the object is within the frame during capture. For example, without this feedback, the user 22 may inadvertently capture images of the ceiling, the sky, or part of his or her hand instead of images of the object itself.
However, it can be challenging for a user to maintain a view of the camera screen during scanning. For example, when the object is low, as depicted in FIG. 2, the user needs to hold the camera level with the object, which can make it difficult for the user to be able to see the screen without crouching. As another example, it may be physically impossible for a user to maintain a view of the screen when there is not enough space to walk all the way around the object, such as when the object is against a wall, as shown in FIG. 2.